1 views collected around this technical thread.
This article explains why queueing systems and rate‑limiting strategies are essential for AIGC platforms, describes the user‑facing product behaviors they produce, outlines design considerations, compares technical options, and provides practical implementation guidance to keep services stable, cost‑effective, and user‑friendly.
Leveraging AI-driven monocular depth estimation, novel view synthesis, and MV‑HEVC encoding, the JD Retail Content R&D team presents an end‑to‑end pipeline that converts 2D video assets into high‑quality immersive 3D spatial videos, introduces the large‑scale StereoV1K dataset, and demonstrates superior performance over existing methods.
This article presents a comprehensive AI‑driven pipeline that converts 2D video into immersive 3D spatial video by leveraging monocular depth estimation, depth‑warping novel view synthesis, a multi‑branch inpainting module, a large‑scale StereoV1K dataset, and efficient MV‑HEVC compression, with results validated at ICME 2025 and deployed in JD Vision services.
This work presents an end‑to‑end AI pipeline that transforms existing monocular 2D videos into immersive 3D spatial streams by combining DINO‑v2‑based depth estimation, multi‑branch view synthesis, and MV‑HEVC encoding, achieving up to 33 % BD‑Rate reduction, 31 % speed gains, state‑of‑the‑art visual quality, and real‑time production suitability, validated on the new StereoV1K benchmark and deployed in JD.Vision’s e‑commerce catalog.
The article presents a series of technical breakthroughs by JD's advertising team that improve the quality and coverage of AI‑generated ad images through a trustworthy multimodal feedback network, introduce a large human‑annotated image dataset, and enhance creative ranking with offline multimodal representations and online architecture optimizations, ultimately achieving more precise and scalable ad personalization.
Taobao’s Content AI team leverages a proprietary multimodal Mixture‑of‑Experts model to automatically generate short‑form videos—extracting highlights from live streams and creating customized product explainers—using two‑stage CLIP/VideoBLIP training, character‑level timestamps, LLM re‑segmentation and OCR masking, now producing over 100 k daily videos with a 12 % approval boost and notable conversion gains.
Ele.me created a domain-specific AIGC image model built from scratch on its own data using the DiT backbone, a three-stage training pipeline (transformer pre-training, prompt alignment, aesthetic fine-tuning), custom T5‑E‑CLIP text and visual encoders, ControlNet for layout control, and evaluated via FID, CLIP scores and a human rubric, enabling automated dish-image generation and UI asset creation for its vertical business.
This article outlines the rapid growth of AIGC technologies, describes key image‑generation and language models, demonstrates step‑by‑step design workflows, explores user‑experience research enhancements, and envisions future business uses while offering practical tips for mastering AI‑generated content.
In 2024 the advertising team deployed an end‑to‑end AIGC pipeline that automatically creates high‑quality ad images, uses the multimodal Reliable Feedback Network and the million‑size RF1M dataset to filter outputs, builds rich offline and online multimodal representations with contrastive and list‑wise learning, and optimizes ranking architecture to deliver scalable, personalized creative selection.
By 2025, AIGC is poised to transform full‑stack development, automating requirement analysis, code generation, testing, and deployment, thereby accelerating productivity while prompting developers to upskill and collaborate with AI rather than fear replacement.
The 2024 advertising team introduced a suite of AI‑driven techniques—including a trustworthy feedback network, a large‑scale human‑annotated dataset, multimodal large language model representations, and online ranking architecture upgrades—to dramatically improve the quality, coverage, and personalization of generated ad creatives.
Taobao’s AIGC pipeline combines a human‑feedback multimodal reward model, audio‑visual joint pre‑training, and Mixture‑of‑Experts distillation to clean data, align outputs with user preferences, and achieve state‑of‑the‑art multimodal LLM performance that drives content cold‑start and conversion gains in e‑commerce.
We introduce an AIGC‑driven content analysis agent for Taobao that automatically builds a fine‑grained feature taxonomy, multimodally labels images, videos and text, mines causal quality features, and deploys auto‑prompt, memory and LoRA‑based training pipelines, achieving over 90% labeling accuracy, 80% labor savings and double‑digit CTR gains while cutting video generation cost to 25% of GPT‑4o.
The article surveys recent AIGC video generation advances for Taobao e‑commerce, detailing lip‑sync models like Wav2Lip and MuseTalk, head‑driven systems such as Hallo and EchoMimic, body‑driven pipelines including AnimateAnyone and Tango, and a four‑stage production workflow that boosts click‑through rates and enables virtual try‑on.
The paper details the design and deployment of an AI‑driven face‑swap platform for the 2025 CCTV Spring Festival Gala, featuring a dual‑model SDXL pipeline with ControlNet and LoRA fine‑tuning, optimized preprocessing and GPU‑specific acceleration to achieve sub‑3‑second latency at over 10 k QPS, supporting scaling, throttling, and multi‑region load balancing, and ultimately serving ten million users and generating hundreds of millions of personalized gala images.
TMIC’s Canvas Engine combines a Konva‑based front‑end canvas framework with AIGC model interfaces to give merchants low‑level, undoable editing tools—such as template loading, intelligent image adjustments, magic eraser, filters, and watermarking—while exposing an SDK and JSON protocol for easy integration and future AI‑driven extensions.
This article examines DeepSeek’s frequent service interruptions, contrasts the inherent reliability challenges of AIGC products with traditional internet applications, and proposes adopting proven isolation, rate‑limiting, and elastic‑scaling techniques to improve AI service availability and user experience.
This article explains what AIGC agents are, outlines their four main characteristics, describes the underlying transformer‑based architecture, dual‑stage learning, probabilistic generation and feedback optimization, and explores their current and future commercial use cases across content creation, knowledge bases, customer service, internal operations, and product design.
In 2024, AI‑generated content transforms fresh‑food e‑commerce by leveraging large language and multimodal models to automatically craft concise product copy, synthesize realistic images and short‑video GIFs, detect and fix visual defects, and build brand‑specific style libraries, paving the way for future voice‑driven storytelling and personalized shopping experiences.
The article outlines OpenAI's five AI maturity levels—from chatbots to organizational AI—examines the challenges Chinese enterprises face when adopting large‑model technologies, and presents the AGILE five‑step framework (Awareness, Gauge, Inception, Ladder, Expansion) together with current best practices and job‑market impacts.